Spectral Graph Attention Network with Fast Eigen-approximation

Variants of Graph Neural Networks (GNNs) for representation learning have been proposed recently and achieved fruitful results in various fields. Among them, Graph Attention Network first employs a self-attention strategy to learn attention weights for each edge in the spatial domain. However, learning the attentions over edges can only focus on the local information of graphs and greatly increases the computational costs. In this paper, we first introduce the attention mechanism in the spectral domain of graphs and present Spectral Graph Attention Network that learns representations for different frequency components regarding weighted filters and graph wavelets bases. In this way, can better capture global patterns of graphs in an efficient manner with much fewer learned parameters than that of. Further, to reduce the computational cost of brought by the eigen-decomposition, we propose a fast approximation variant -. We thoroughly evaluate the performance of and - in semi-supervised node classification tasks and verify the effectiveness of the learned attentions in spectral domain.